Hello!

I'm Ajinkya Talekar,
a Software Engineer
based in New York.

About Me.

I am a third-year student at the University at Buffalo, majoring in Computer Science with a minor in Mathematics.

I currently work as a Software Engineer Intern at SteinnLabs, and part-time as a Computer Science Teaching Assistant at the University at Buffalo. These positions have helped me immensely in developing my technical, communication and presentation skills.

Languages

  • Python
  • JavaScript, TypeScript
  • Java
  • PHP
  • SQL

Frameworks & Libraries

  • React
  • React Native
  • FastAPI
  • Django
  • LangChain

Tools & Technologies

  • AWS
  • Git
  • Docker
  • PostgreSQL
  • TensorFlow
  • Supabase
  • WatermelonDB

Experience.

1. Software Engineer Intern - SteinnLabs

• Architected and deployed a full-stack appointment scheduling system using React and PostgreSQL, implementing real-time updates and secure authentication through Supabase.
• Engineered RESTful APIs using FastAPI and optimized database queries, reducing endpoint latency by 20% and improving application performance across 1000+ daily user interactions.
• Designed and implemented AWS infrastructure utilizing S3 buckets for data storage and processing pipelines, enabling scalable data visualization features for environmental metrics.

Tech: React, React Native, FastAPI, PostgreSQL, Supabase, AWS, Git

2. Teaching Assistant - University at Buffalo

• Facilitated learning of 600+ students in Discrete Structures for Data Structures and Algorithms.
• Demonstrated strong communication and leadership skills by managing and teaching class of over 40 students weekly and providing office hours for individual help.
• Produced code and slides, and assessed tests to support diverse learning styles and improve student engagement.

Projects.

1. Autovid - Full Stack Python Package

Autovid is an open-source python package for automated video generation. It uses various text-to-speech, video, and audio manipulation libraries in Python to produce high-quality videos with minimal user input.
Autovid's new beta version also features a GUI made using Django, allowing the user to easily create videos without the hassle of command line inputs.

Tech: Python, Django, Git
Check it out on GitHub.

2. SwiftGesture - Hand Gesture Classification Pipeline

SwiftGesture is a pipeline for rapidly training fast real-time models for classifying custom hand gestures. It produces reliable models with just 3 input images per gesture being trained.
A model trained using SwiftGesture on just 80 images was able to accurately classify ALL the letters of the American sign language with 95% accuracy.

Tech: Python, Tensorflow, MediaPipe
Check it out on GitHub.

3. Wordle Puzzle Solver

Programmed an algorithm to solve the word guessing puzzle (Wordle) by The New York Times efficiently using JavaScript.
Designed an attractive and user-friendly UI, matching the theme of the official puzzle site.

Tech: JavaScript
Run it here.